380
British Journal of Industrial Medicine 1990;47:380-383
Occupational risk factors for upper respiratory tract and upper digestive tract cancers JM Haguenoer, S Cordier, C Morel, JL Lefebvre, D Hemon Abstract A case-control study was conducted to investigate occupational risk factors for upper respiratory and digestive tract cancers (nose, lips, buccal cavity, pharynx, and larynx) in the north of France. Two hundred and eighty three cases of histologically confirmed cancer in men treated during the first semester of 1983 at the Regional Cancer Institute were included in the study. Two controls per case were chosen from patients in the medical wards of the local general hospitals in the same geographical area as the case, who did not have cancer. Controls were individually matched for sex, age- (+ 18 months), ethnic group, area of residence, and smoking and alcohol drinking history. All subjects were questioned about occupations in which they had worked for at least 15 -years. Odds ratios for major occupational categories were estimated using conditional logistic regression for matched samples. Significant associations were found between wood work and nasal cancer (four cases, no control) and farming and lip cancer (odds ratio 5 3, 95% confidence interval 1-126-8). Pharyngeal cancer was associated with the textile industry (OR 2-4, 95% CI 1-0-5-7) and the building industry (OR 2-0, 95% CI 1 1-3-9). Coal miners showed a threefold excess risk for cancer of the lip (4 cases, no control), buccal cavity (OR 3-5, 95% CI 1-1-11-8), and larynx (OR 3-2, 95% CI 1-1-9-7).
Mortality from upper respiratory and digestive tract (URDT; nose and sinus, lips, buccal cavity, pharynx, and larynx) cancers in men is higher in France than in all other western countries (27-4 per 100 000 in 1977 compared with 12 5 in Italy which ranks second, and IMT, Region du Nord, Facultk de Medecine, 59045 Lille Cedex, France JM Haguenoer, C Morel INSERM U 170, 94807 Villejuif Cedex S Cordier, D Hemon Centre de Lutte contre le Cancer du Nord de la France, 59020 Lille Cedex JL Lefebvre
3-5 in Sweden), and has increased during the past 20
years.' Comparisons of incidence rates from the available registries show the same trends.2 Within France, mortality from URDT cancers in men is higher in the north and in Brittany.' The same observations for oesophageal cancer prompted a study which showed the major part played by alcohol and tobacco consumption in the aetiology of this cancer.4 No doubt the high incidence of URDT cancers in the north of France might also be partly explained by regional alcohol drinking and smoking habits. Considering the high level of industrial activity in this area, however, a case-control study was undertaken to investigate the possible role of occupational factors in URDT cancers, taking into account alcohol and tobacco consumption.
Methods Two hundred and eighty three men treated during the first semester of 1983 for histologically confirmed primary URDT cancer at the Regional Cancer Institute were included in our study. This institute is the only centre of its type in the north of France (Nord-Pas de Calais region), and treats over two thirds of URDT cancer patients diagnosed in the region. Examination of the statistics of admission showed no bias for selection on socioprofessional characteristics compared with the region as a whole. Two hundred and thirty four patients (82 6% ) had been diagnosed less than one year before inclusion in the study. For each case two men were chosen as controls from patients in the wards of internal medicine of local general hospitals in the same geographical area as the case, who did not have cancer. Controls were individually matched for age (± 18 months), country of origin (France, southern Europe, eastern Europe, north Africa), area of residence, and smoking and alcohol drinking history. Tobacco consumption was expressed as cigarette equivalents (one pipe = two cigars = four cigarettes). Matching was done separately on class of average daily tobacco consumption (non-smoker or ex-smoker for more than five years, one to nine cigarettes a day, 10-19 cigarettes a day, ) 20 cigarettes a day) and number of years smoked (± five years). Alcohol drinkers were
Occupational risk factors for upper respiratory tract and upper digestive tract cancers
381
matched according to four groups of average daily Table 2 Distribution of cases and controls according to consumption (occasional drinkers, < 44 g a day, tobacco and alcohol consumption 44-88 g a day, > 88 g a day; equivalents of pure Cases (%) Controls (%) alcohol, 1 1 wine = 88 g, 1 1 beer = 44 g, 1 measure (n- 283) (n = 566) spirits= 15 g). Potential controls were first interviewed about their smoking and alcohol drinking Tobacco consumption (pack-years) habits; if matching was not achieved the interview 0 9 (3-2) 17 (3-0) 1-19 52 (18-4) 95 (16-8) was not completed. 20-39 134 (47-5) 268 (47-5) The causes for entering hospital among the 566 40-59 45 (16-0) 110 (19 5) 43 (15-2) 76 (13 4) controls were cardiovascular diseases (25%), meta- > 60 bolic diseases (17o0), infectious diseases (160o), Alcohol consumption (glday) digestive problems (1500), neurological disorders Occasional drinker 42 (14 8) 85 (15-0) < 44 48 (170) 95 (168) (8%,), skeletomuscular diseases (80oo), and miscel- 44-88 96 (33 9) 192 (33-9) laneous diseases (11 "0). > 88 97 (34 3) 194 (34-3) Cases and controls were interviewed while in hospital by the same person (CM) about their accounted for more than 80°, of the cases. Ninety smoking and alcohol drinking history and about the three per cent of the cancers were squamous cell occupations in which they had worked for at least 15 carcinomas. years. Subjects who had not worked in the same type Close matching was achieved between patients of job for at least 15 years were not included in the with cancer and controls for country of origin (920o study. This led to the exclusion of half the subjects. of patients were French), area of residence, age at Types of occupation were classified into 11 interview (mean age 58 3 years), and tobacco and groups-namely, coal mining, the building industry, alcohol consumption (table 2). Tobacco consumpmetal work and mechanics, agriculture, road trans- tion was slightly lower, and alcohol consumption was port, shipping, the textile industry, wood work, road much higher than in other case-control studies on works, service industries, and others. Only two oral cavity, pharyngeal, or laryngeal cancers pubpeople had held two different jobs for more than 15 lished in the United States and Canada. Table 3 shows the odds ratio estimates for 11 years each and they were assigned to both groups in which they had worked. Odds ratios for having occupational groups for all sites of cancer combined. worked in each occupational group were calculated These ratios compare the risk of disease for people using the Mantel-Haenszel estimation and con- who worked in these occupations for more than 15 years with the risk of disease for people who never ditional logistic regression for matched samples.' worked in these occupations or those who worked for less than 15 years. Relative risks were significantly Results Table 1 describes the different sites and histological greater than one among coal miners, workers in the types of URDT cancers included in the study. building industry, and wood workers. Relative risk of Cancers of the pharynx, buccal cavity, and larynx URDT cancers for service industries was significantly lower than one. Table I Distribution of cases by site and histological type Analyses related to specific sites of cancer are for URDT cancers presented in table 4. Among 14 cases of nose and sinus cancer, four were wood workers. Three had Cases (°,) (n = 283j
Table 3 Estimates of odds ratios for different occupational groups for all sites of cancer combined
Site of cancer
Nose, sinus Lips Buccal cavity Pharynx: Oropharynx Hypopharynx Larynx: Supralarynx Endolarynx Other
Multiple sites T
otal
14 16 64 114
(4 9) (5 7) (22 6) (40 3)
950 Occupational group
Cases
54 (19-1)
Mining Building industry Metal work,
48 53 44
9 (3 2) 12 (4 2)
Agriculture Road transport Shipping and
78
36 12 37 283 (100-0)
Histological type for non-multiple sites 252 (93-0) Squamous cell 7 (2 6) Glandular 12 (4-4) cancer Non-epithelial 271 (100-0) Total
ratio
Confidence interval
52 59 83
2.6*
1-6-4 3
2 0* 1.1
0 7-1-6
18 17 14
42 37 32
09 09 0-8
0-5-1 5 0-5-1-7 04-1 7
17
27 7 27 179 21
1 3
0 7-2-5 1 2-10 3 0-5-1-9 0-3-0 6
(n
Controls
2831 !n = 566)
Odds
13-3.0
mechanics
seamen
Textile industry Wood work Road works Service industries
Others
10 13 45 7
3.5* 10 0 4* 0-7
*Odds ratio significantly different from one (p < 0-05).
03-1-6
Haguenoer, Cordier, Morel, Lefebvre, Hemon
382
Table 4 Estimation of odds ratios by occupational group for each site of cancer (No of exposed cases/No of exposed controls in parentheses)
Mining Building industry Metal work, mechanics Agriculture Road transport Shipping and seamen TIextile industry Wood work Road works Service industries Others
Nose sinus (n=14)
Lips (n= 16)
6 0 (3/1) 0 (0/4) - (3/0) 1.3 (2/3) 1 0 (1/2)
-* (4/0) 2-4 (3/3) 0 (0/5) 5 3* (6/3) 0 (0/2)
0(0/2)
0 (0/1) -* (4/0)
0(0/2) 0 2 (1/11) 0 (0/2)
0(0/1) 0 (0/2) 0 (0/2) 2 7 (2/2) 0.1* (1/11) 0 (0/1)
Buccal cavity (n=64)
Pharynx (n= 114)
Larynx (n=54)
3 5* (11/11) 1 9 (12/14) 1.2 (13/22) 0.3 (1/7) 0 8 (4/10) 0 5 (3/9) 0 7 (2/5) 1 3 (2/3) 2 2 (5/5) 0 5 (11/39) 0 7 (1/3)
1-4 (17/28) 2-0* (21/23) 0-7 (16/42) 0.5 (4/15) 0-8 (6/15) 1-0 (5/10) 2-4* (12/11) -(4/1) 1-8 (5/6) 0.5* (20/67) 1-0 (5/10)
3 2* (l1/10)
1 5 (8/11) 1 8 (7/8) 1-0 (5/10) 2 2 (6/6) 1 8 (6/8) 0 3 (1/6) 0 (0/1) 0-2(1/10) 0.4* (8/34) 0 5 (1/4)
Multiple sites (n= 12)
All sites (n= 283)
-(2/1) -* (8/2) 1 6 (3/4) 0(0/3) 0 (0/1) 0(0/1) 0 (0/1) -(0/0) 0(0/2) 0 (0/8) 0 (0/1)
2-6* (48/52) 2 0* (53/59) 1.1 (44/83) 0.9 (18/42) 0-9 (17/37) 0 8 (14/31) 1 3 (17/27) 3.5* (10/7) 1 0 (13/27) 0.4* (45/179) 0-7 (7/21)
*Odds ratio significantly different from one (p < 0-05),-indeterminate (matched odds ratio could not be computed).
cancer of the ethmoids and one had cancer of the namely, spinners and winders (4 cases, 5 controls), maxillary sinus. Relative risk for cancer of the lip was weavers (2 cases, 3 controls), pressers (1 case, no significantly greater than one among coal miners and control), maintenance workers (2 cases, 1 control), among farmers. The only significant excess risk of and unknown occupations (3 cases, 2 controls). cancer of the buccal cavity was among coal miners Excess risk was also found in workers in the building who were also at higher relative risk for laryngeal industry. An excess of oral and pharyngeal cancer among cancer. A significant relative risk of pharyngeal cancer was shown among workers in the building textile workers in England has been reported previously,8 but this was not confirmed by a later industry and in the textile industry. study.9 Leather work was associated with cancer of the buccal cavity, pharynx, and larynx,'0 and Elwood and colleagues reported a moderately increased Discussion It is appreciated that restricting choice of subjects to relative risk for these cancers in men with a history of those who had held one job for at least 15 years working in the logging and mining industries. " More introduced a selection; it was the same, however, for recently, a study carried out in France on workers both cases and controls. Such a selection precluded producing manmade mineral fibres showed them to the study of short term exposures of employees, and have an excess risk for cancer of the pharynx, the of more recent industrial activities. In computing the buccal cavity, and the larynx.'2 Work in the building industry involves various odds ratio for a particular occupational group the reference category could include men who had types of exposure, especially to cement dust and worked in the same group for less than 15 years, or in insulation material, which might influence the other occupational groups potentially at risk, and this occurrence ofpharyngeal cancer in workers from this could bias the results. This seldom happened, industry. In our study 15 cases and 11 controls however, as employment in the region is stable, and working in the building industry had jobs involving the number of years spent in the reported occupa- heavy dust exposure (masons, tile layers, plasterers), tions covered most of the working life of both cases compared with less dusty jobs (painters, electricians, and their matched controls. Furthermore, the choice plumbers) in which six cases and 12 controls had of this reference category tends to underestimate the worked. relative risks. LARYNX
Occupational risk factors for iaryngeal cancer have The associations found in the nose and lips are been extensively studied. Asbestos, nickel refining, already established. For instance, the association wood dust, mustard gas manufacturing, isopropyl between exposure to wood dust and nasal cancer has alcohol, leather work, and sulphuric acid have all been reported6 and lip cancer has been shown to be been associated with an excess risk of laryngeal more frequent among people working outdoors, cancer.'3 More recently, an increased relative risk was found among farmers and textile processors. '4 Olsen especially farmers.7 and colleagues in Denmark reported high relative risks among workers exposed to dust, drivers, people PHARYNX In the present study an excess risk of pharyngeal working in the cement industries and port services,'5 cancer was seen among workers in the textile and among workers exposed to welding fumes.'6 A industry and occurred in a spread of subgroups- study in Connecticut found a significantly raised NOSE AND LIPS
Occupational risk factors for upper respiratory tract and upper digestive tract cancers
383
relative risk for machinists,'7 and a non-specific We thank Laurence Mandereau for technical help association was found between laryngeal cancer and and Evelyne Przybilski for typing the manuscript. work as a non-construction labourer.'8 Finally, a 1 Moulin JJ, Mur JM, Cavelier C. Epidemiologie comparee, en mortality study of workers employed in a refractory Europe, des cancers lies au tabac (poumon, larynx, pharynx, brick plant showed an increased standardised morcavite buccale). Bull Cancer 1985;72:155-8. 2 Waterhouse J, Muir C, Shanmugaratham K, Powell J, eds. tality ratio for laryngeal cancer among workers Cancer incidence infive continents. Vol IV. Lyon: International exposed to silica dust suggesting a role for exposure Agency for Research on Cancer, 1982. (IARC sci publ No 42.) 3 Rezvani A, Doyon F, Flamant R, eds. Atlas de la mortalite par to silica in the aetiology of this cancer.'9 In the cancer en France (1971-1978). Paris: INSERM, 1986. present study excess of laryngeal cancer occurred 4 Tuyns AJ, Pequignot G, Jensen OM. Le cancer de l'oesophage en Ile-et-Vilaine en fonction des niveaux de consommation among coal miners. This could be attributed to d'alcool et de tabac. Des risques qui se multiplient. Bull exposures to silica dust and oil mists. This point is Cancer 1977;64:45-60. further below. discussed 5 Breslow NE, Day NE. Statistical methods in cancer research. Vol 6 COAL MINERS
In our study coal miners were the only group in which three sites of URDT cancers appeared in excess. These were cancers of the lip, buccal cavity, and larynx. Working in coal mines involves specific exposures related to different habits of life style and presence of dusts, fumes, or vapours inside the galleries. Coal miners have reported making frequent use of chewing tobacco, a major risk factor for cancer of the buccal cavity.20 We were unable, however, to quantify this association. In the coal mines of the Nord-Pas de Calais region no exhaust fumes were present as no combustion engines were used. Industrial hygienists have recently been concerned with the presence of mineral oils used in hydraulic operations and mineral oils have been linked to laryngeal cancer. In addition to the raised relative risk among machinists already mentioned'7 a study of subsequent primary tumours among 288 men with scrotal cancers, most likely induced by cutting oils, found a significant excess of tumours of the larynx, lip, bronchus, stomach, and skin.2' Several mortality studies have been conducted among coal miners. An excess of stomach cancer has been found, but no associations with the sites studied here.22 This might be due to competing risks which were stronger in earlier decades. It is also possible that we were able to show such associations because of the high proportion of coal miners (up to 200%o of the workforce in the 1950s in Nord-Pas de Calais) in a community where patients with URDT cancers were easily accessible because ofthe high background rate, due probably to alcohol drinking habits in the region.
7
8 9 10 11
12
13 14 15
16 17
18 19
20 21
1. The analysis of case-control studies. Lyon: International Agency for Research on Cancer, 1980. (IARC sci publ No 32.) International Agency for Research on Cancer. Monographs on the evaluation of the carcinogenic risk of chemicals to humans. Vol 25. Wood, leather and some associated industries. Lyon: IARC, 1981. Lindqvist C. Risk factors in lip cancer: a questionnaire survey. Am J Epidemiol 1979;109:521-30. Moss E, Lee WR. Occurrence of oral and pharyngeal cancers in textile workers. Br J Ind Med 1974;31:224-32. Whitaker CJ, Moss E, Lee WR, et al. Oral and pharyngeal cancer in the north-west and west Yorkshire regions of England, and occupation. Br J Ind Med 1979;36:292-8. Decoufle P. Cancer risks associated with employment in the leather and leather products industry. Arch Environ Health 1979;34:33-7. Elwood JM, Pearson JCG, Skippen DH, et al.Alcohol, smoking, social and occupational factors in the aetiology of cancer of the oral cavity, pharynx and larynx. Int J Cancer 1984;34: 603-12. Moulin JJ, Mur JM, Wild P, et al. Oral cavity and laryngeal cancers among man-made mineral fiber production workers. Scand J Work Environ Health 1986;12:27-31. Rothman KJ, Cann CI, Flanders D, et al. Epidemiology of laryngeal cancer. Epidemiol Rev 1980;2:195-209. Flanders WD, Cann CI, Rothman KJ. Work-related risk factors for laryngeal cancer. Am J Epidemiol 1984;119:23-32. Olsen J, Sabroe S. Occupational causes of laryngeal cancer. J Epidemiol Community Health 1984;38:117-21. Olsen J, Sabroe S, Lajer M. Welding and cancer of the larynx: a case-control study. Eur J Cancer Clin Oncol 1984;20: 639-43. Zagraniski RT, Kelsey J, Walter SD. Occupational risk factors for laryngeal carcinoma: Connecticut, 1975-1980. Am J Epidemiol 1986;124:67-76. Brownson RC, Chang JC. Exposure to alcohol and tobacco and the risk of laryngeal cancer. Arch Environ Health 1987;42: 192-6. Puntoni R, Goldsmith DF, Vercelli M, et al. A cohort study of workers employed in a refractory brick plant. Scand J Work Environ Health 1987;2:162. International Agency for Research on Cancer. Monographs on the evaluation of carcinogenic risk of chemicals to humans. Vol 37. Tobacco habits other than smoking. Lyon: IARC, 1985. Waldron HA. The carcinogenicity of oil mist. Br J Cancer
1975;32:256-7.
22 Simonato L, Saracci R. Cancer occupational. In: Encyclopedia of occupational health and safety. Vol 1. Geneva: International Labour Office, 1983:369-75.
Accepted 13 November 1989
British Journal of Industrial Medicine 1990;47:380-383
Occupational risk factors for upper respiratory tract and upper digestive tract cancers JM Haguenoer, S Cordier, C Morel, JL Lefebvre, D Hemon Abstract A case-control study was conducted to investigate occupational risk factors for upper respiratory and digestive tract cancers (nose, lips, buccal cavity, pharynx, and larynx) in the north of France. Two hundred and eighty three cases of histologically confirmed cancer in men treated during the first semester of 1983 at the Regional Cancer Institute were included in the study. Two controls per case were chosen from patients in the medical wards of the local general hospitals in the same geographical area as the case, who did not have cancer. Controls were individually matched for sex, age- (+ 18 months), ethnic group, area of residence, and smoking and alcohol drinking history. All subjects were questioned about occupations in which they had worked for at least 15 -years. Odds ratios for major occupational categories were estimated using conditional logistic regression for matched samples. Significant associations were found between wood work and nasal cancer (four cases, no control) and farming and lip cancer (odds ratio 5 3, 95% confidence interval 1-126-8). Pharyngeal cancer was associated with the textile industry (OR 2-4, 95% CI 1-0-5-7) and the building industry (OR 2-0, 95% CI 1 1-3-9). Coal miners showed a threefold excess risk for cancer of the lip (4 cases, no control), buccal cavity (OR 3-5, 95% CI 1-1-11-8), and larynx (OR 3-2, 95% CI 1-1-9-7).
Mortality from upper respiratory and digestive tract (URDT; nose and sinus, lips, buccal cavity, pharynx, and larynx) cancers in men is higher in France than in all other western countries (27-4 per 100 000 in 1977 compared with 12 5 in Italy which ranks second, and IMT, Region du Nord, Facultk de Medecine, 59045 Lille Cedex, France JM Haguenoer, C Morel INSERM U 170, 94807 Villejuif Cedex S Cordier, D Hemon Centre de Lutte contre le Cancer du Nord de la France, 59020 Lille Cedex JL Lefebvre
3-5 in Sweden), and has increased during the past 20
years.' Comparisons of incidence rates from the available registries show the same trends.2 Within France, mortality from URDT cancers in men is higher in the north and in Brittany.' The same observations for oesophageal cancer prompted a study which showed the major part played by alcohol and tobacco consumption in the aetiology of this cancer.4 No doubt the high incidence of URDT cancers in the north of France might also be partly explained by regional alcohol drinking and smoking habits. Considering the high level of industrial activity in this area, however, a case-control study was undertaken to investigate the possible role of occupational factors in URDT cancers, taking into account alcohol and tobacco consumption.
Methods Two hundred and eighty three men treated during the first semester of 1983 for histologically confirmed primary URDT cancer at the Regional Cancer Institute were included in our study. This institute is the only centre of its type in the north of France (Nord-Pas de Calais region), and treats over two thirds of URDT cancer patients diagnosed in the region. Examination of the statistics of admission showed no bias for selection on socioprofessional characteristics compared with the region as a whole. Two hundred and thirty four patients (82 6% ) had been diagnosed less than one year before inclusion in the study. For each case two men were chosen as controls from patients in the wards of internal medicine of local general hospitals in the same geographical area as the case, who did not have cancer. Controls were individually matched for age (± 18 months), country of origin (France, southern Europe, eastern Europe, north Africa), area of residence, and smoking and alcohol drinking history. Tobacco consumption was expressed as cigarette equivalents (one pipe = two cigars = four cigarettes). Matching was done separately on class of average daily tobacco consumption (non-smoker or ex-smoker for more than five years, one to nine cigarettes a day, 10-19 cigarettes a day, ) 20 cigarettes a day) and number of years smoked (± five years). Alcohol drinkers were
Occupational risk factors for upper respiratory tract and upper digestive tract cancers
381
matched according to four groups of average daily Table 2 Distribution of cases and controls according to consumption (occasional drinkers, < 44 g a day, tobacco and alcohol consumption 44-88 g a day, > 88 g a day; equivalents of pure Cases (%) Controls (%) alcohol, 1 1 wine = 88 g, 1 1 beer = 44 g, 1 measure (n- 283) (n = 566) spirits= 15 g). Potential controls were first interviewed about their smoking and alcohol drinking Tobacco consumption (pack-years) habits; if matching was not achieved the interview 0 9 (3-2) 17 (3-0) 1-19 52 (18-4) 95 (16-8) was not completed. 20-39 134 (47-5) 268 (47-5) The causes for entering hospital among the 566 40-59 45 (16-0) 110 (19 5) 43 (15-2) 76 (13 4) controls were cardiovascular diseases (25%), meta- > 60 bolic diseases (17o0), infectious diseases (160o), Alcohol consumption (glday) digestive problems (1500), neurological disorders Occasional drinker 42 (14 8) 85 (15-0) < 44 48 (170) 95 (168) (8%,), skeletomuscular diseases (80oo), and miscel- 44-88 96 (33 9) 192 (33-9) laneous diseases (11 "0). > 88 97 (34 3) 194 (34-3) Cases and controls were interviewed while in hospital by the same person (CM) about their accounted for more than 80°, of the cases. Ninety smoking and alcohol drinking history and about the three per cent of the cancers were squamous cell occupations in which they had worked for at least 15 carcinomas. years. Subjects who had not worked in the same type Close matching was achieved between patients of job for at least 15 years were not included in the with cancer and controls for country of origin (920o study. This led to the exclusion of half the subjects. of patients were French), area of residence, age at Types of occupation were classified into 11 interview (mean age 58 3 years), and tobacco and groups-namely, coal mining, the building industry, alcohol consumption (table 2). Tobacco consumpmetal work and mechanics, agriculture, road trans- tion was slightly lower, and alcohol consumption was port, shipping, the textile industry, wood work, road much higher than in other case-control studies on works, service industries, and others. Only two oral cavity, pharyngeal, or laryngeal cancers pubpeople had held two different jobs for more than 15 lished in the United States and Canada. Table 3 shows the odds ratio estimates for 11 years each and they were assigned to both groups in which they had worked. Odds ratios for having occupational groups for all sites of cancer combined. worked in each occupational group were calculated These ratios compare the risk of disease for people using the Mantel-Haenszel estimation and con- who worked in these occupations for more than 15 years with the risk of disease for people who never ditional logistic regression for matched samples.' worked in these occupations or those who worked for less than 15 years. Relative risks were significantly Results Table 1 describes the different sites and histological greater than one among coal miners, workers in the types of URDT cancers included in the study. building industry, and wood workers. Relative risk of Cancers of the pharynx, buccal cavity, and larynx URDT cancers for service industries was significantly lower than one. Table I Distribution of cases by site and histological type Analyses related to specific sites of cancer are for URDT cancers presented in table 4. Among 14 cases of nose and sinus cancer, four were wood workers. Three had Cases (°,) (n = 283j
Table 3 Estimates of odds ratios for different occupational groups for all sites of cancer combined
Site of cancer
Nose, sinus Lips Buccal cavity Pharynx: Oropharynx Hypopharynx Larynx: Supralarynx Endolarynx Other
Multiple sites T
otal
14 16 64 114
(4 9) (5 7) (22 6) (40 3)
950 Occupational group
Cases
54 (19-1)
Mining Building industry Metal work,
48 53 44
9 (3 2) 12 (4 2)
Agriculture Road transport Shipping and
78
36 12 37 283 (100-0)
Histological type for non-multiple sites 252 (93-0) Squamous cell 7 (2 6) Glandular 12 (4-4) cancer Non-epithelial 271 (100-0) Total
ratio
Confidence interval
52 59 83
2.6*
1-6-4 3
2 0* 1.1
0 7-1-6
18 17 14
42 37 32
09 09 0-8
0-5-1 5 0-5-1-7 04-1 7
17
27 7 27 179 21
1 3
0 7-2-5 1 2-10 3 0-5-1-9 0-3-0 6
(n
Controls
2831 !n = 566)
Odds
13-3.0
mechanics
seamen
Textile industry Wood work Road works Service industries
Others
10 13 45 7
3.5* 10 0 4* 0-7
*Odds ratio significantly different from one (p < 0-05).
03-1-6
Haguenoer, Cordier, Morel, Lefebvre, Hemon
382
Table 4 Estimation of odds ratios by occupational group for each site of cancer (No of exposed cases/No of exposed controls in parentheses)
Mining Building industry Metal work, mechanics Agriculture Road transport Shipping and seamen TIextile industry Wood work Road works Service industries Others
Nose sinus (n=14)
Lips (n= 16)
6 0 (3/1) 0 (0/4) - (3/0) 1.3 (2/3) 1 0 (1/2)
-* (4/0) 2-4 (3/3) 0 (0/5) 5 3* (6/3) 0 (0/2)
0(0/2)
0 (0/1) -* (4/0)
0(0/2) 0 2 (1/11) 0 (0/2)
0(0/1) 0 (0/2) 0 (0/2) 2 7 (2/2) 0.1* (1/11) 0 (0/1)
Buccal cavity (n=64)
Pharynx (n= 114)
Larynx (n=54)
3 5* (11/11) 1 9 (12/14) 1.2 (13/22) 0.3 (1/7) 0 8 (4/10) 0 5 (3/9) 0 7 (2/5) 1 3 (2/3) 2 2 (5/5) 0 5 (11/39) 0 7 (1/3)
1-4 (17/28) 2-0* (21/23) 0-7 (16/42) 0.5 (4/15) 0-8 (6/15) 1-0 (5/10) 2-4* (12/11) -(4/1) 1-8 (5/6) 0.5* (20/67) 1-0 (5/10)
3 2* (l1/10)
1 5 (8/11) 1 8 (7/8) 1-0 (5/10) 2 2 (6/6) 1 8 (6/8) 0 3 (1/6) 0 (0/1) 0-2(1/10) 0.4* (8/34) 0 5 (1/4)
Multiple sites (n= 12)
All sites (n= 283)
-(2/1) -* (8/2) 1 6 (3/4) 0(0/3) 0 (0/1) 0(0/1) 0 (0/1) -(0/0) 0(0/2) 0 (0/8) 0 (0/1)
2-6* (48/52) 2 0* (53/59) 1.1 (44/83) 0.9 (18/42) 0-9 (17/37) 0 8 (14/31) 1 3 (17/27) 3.5* (10/7) 1 0 (13/27) 0.4* (45/179) 0-7 (7/21)
*Odds ratio significantly different from one (p < 0-05),-indeterminate (matched odds ratio could not be computed).
cancer of the ethmoids and one had cancer of the namely, spinners and winders (4 cases, 5 controls), maxillary sinus. Relative risk for cancer of the lip was weavers (2 cases, 3 controls), pressers (1 case, no significantly greater than one among coal miners and control), maintenance workers (2 cases, 1 control), among farmers. The only significant excess risk of and unknown occupations (3 cases, 2 controls). cancer of the buccal cavity was among coal miners Excess risk was also found in workers in the building who were also at higher relative risk for laryngeal industry. An excess of oral and pharyngeal cancer among cancer. A significant relative risk of pharyngeal cancer was shown among workers in the building textile workers in England has been reported previously,8 but this was not confirmed by a later industry and in the textile industry. study.9 Leather work was associated with cancer of the buccal cavity, pharynx, and larynx,'0 and Elwood and colleagues reported a moderately increased Discussion It is appreciated that restricting choice of subjects to relative risk for these cancers in men with a history of those who had held one job for at least 15 years working in the logging and mining industries. " More introduced a selection; it was the same, however, for recently, a study carried out in France on workers both cases and controls. Such a selection precluded producing manmade mineral fibres showed them to the study of short term exposures of employees, and have an excess risk for cancer of the pharynx, the of more recent industrial activities. In computing the buccal cavity, and the larynx.'2 Work in the building industry involves various odds ratio for a particular occupational group the reference category could include men who had types of exposure, especially to cement dust and worked in the same group for less than 15 years, or in insulation material, which might influence the other occupational groups potentially at risk, and this occurrence ofpharyngeal cancer in workers from this could bias the results. This seldom happened, industry. In our study 15 cases and 11 controls however, as employment in the region is stable, and working in the building industry had jobs involving the number of years spent in the reported occupa- heavy dust exposure (masons, tile layers, plasterers), tions covered most of the working life of both cases compared with less dusty jobs (painters, electricians, and their matched controls. Furthermore, the choice plumbers) in which six cases and 12 controls had of this reference category tends to underestimate the worked. relative risks. LARYNX
Occupational risk factors for iaryngeal cancer have The associations found in the nose and lips are been extensively studied. Asbestos, nickel refining, already established. For instance, the association wood dust, mustard gas manufacturing, isopropyl between exposure to wood dust and nasal cancer has alcohol, leather work, and sulphuric acid have all been reported6 and lip cancer has been shown to be been associated with an excess risk of laryngeal more frequent among people working outdoors, cancer.'3 More recently, an increased relative risk was found among farmers and textile processors. '4 Olsen especially farmers.7 and colleagues in Denmark reported high relative risks among workers exposed to dust, drivers, people PHARYNX In the present study an excess risk of pharyngeal working in the cement industries and port services,'5 cancer was seen among workers in the textile and among workers exposed to welding fumes.'6 A industry and occurred in a spread of subgroups- study in Connecticut found a significantly raised NOSE AND LIPS
Occupational risk factors for upper respiratory tract and upper digestive tract cancers
383
relative risk for machinists,'7 and a non-specific We thank Laurence Mandereau for technical help association was found between laryngeal cancer and and Evelyne Przybilski for typing the manuscript. work as a non-construction labourer.'8 Finally, a 1 Moulin JJ, Mur JM, Cavelier C. Epidemiologie comparee, en mortality study of workers employed in a refractory Europe, des cancers lies au tabac (poumon, larynx, pharynx, brick plant showed an increased standardised morcavite buccale). Bull Cancer 1985;72:155-8. 2 Waterhouse J, Muir C, Shanmugaratham K, Powell J, eds. tality ratio for laryngeal cancer among workers Cancer incidence infive continents. Vol IV. Lyon: International exposed to silica dust suggesting a role for exposure Agency for Research on Cancer, 1982. (IARC sci publ No 42.) 3 Rezvani A, Doyon F, Flamant R, eds. Atlas de la mortalite par to silica in the aetiology of this cancer.'9 In the cancer en France (1971-1978). Paris: INSERM, 1986. present study excess of laryngeal cancer occurred 4 Tuyns AJ, Pequignot G, Jensen OM. Le cancer de l'oesophage en Ile-et-Vilaine en fonction des niveaux de consommation among coal miners. This could be attributed to d'alcool et de tabac. Des risques qui se multiplient. Bull exposures to silica dust and oil mists. This point is Cancer 1977;64:45-60. further below. discussed 5 Breslow NE, Day NE. Statistical methods in cancer research. Vol 6 COAL MINERS
In our study coal miners were the only group in which three sites of URDT cancers appeared in excess. These were cancers of the lip, buccal cavity, and larynx. Working in coal mines involves specific exposures related to different habits of life style and presence of dusts, fumes, or vapours inside the galleries. Coal miners have reported making frequent use of chewing tobacco, a major risk factor for cancer of the buccal cavity.20 We were unable, however, to quantify this association. In the coal mines of the Nord-Pas de Calais region no exhaust fumes were present as no combustion engines were used. Industrial hygienists have recently been concerned with the presence of mineral oils used in hydraulic operations and mineral oils have been linked to laryngeal cancer. In addition to the raised relative risk among machinists already mentioned'7 a study of subsequent primary tumours among 288 men with scrotal cancers, most likely induced by cutting oils, found a significant excess of tumours of the larynx, lip, bronchus, stomach, and skin.2' Several mortality studies have been conducted among coal miners. An excess of stomach cancer has been found, but no associations with the sites studied here.22 This might be due to competing risks which were stronger in earlier decades. It is also possible that we were able to show such associations because of the high proportion of coal miners (up to 200%o of the workforce in the 1950s in Nord-Pas de Calais) in a community where patients with URDT cancers were easily accessible because ofthe high background rate, due probably to alcohol drinking habits in the region.
7
8 9 10 11
12
13 14 15
16 17
18 19
20 21
1. The analysis of case-control studies. Lyon: International Agency for Research on Cancer, 1980. (IARC sci publ No 32.) International Agency for Research on Cancer. Monographs on the evaluation of the carcinogenic risk of chemicals to humans. Vol 25. Wood, leather and some associated industries. Lyon: IARC, 1981. Lindqvist C. Risk factors in lip cancer: a questionnaire survey. Am J Epidemiol 1979;109:521-30. Moss E, Lee WR. Occurrence of oral and pharyngeal cancers in textile workers. Br J Ind Med 1974;31:224-32. Whitaker CJ, Moss E, Lee WR, et al. Oral and pharyngeal cancer in the north-west and west Yorkshire regions of England, and occupation. Br J Ind Med 1979;36:292-8. Decoufle P. Cancer risks associated with employment in the leather and leather products industry. Arch Environ Health 1979;34:33-7. Elwood JM, Pearson JCG, Skippen DH, et al.Alcohol, smoking, social and occupational factors in the aetiology of cancer of the oral cavity, pharynx and larynx. Int J Cancer 1984;34: 603-12. Moulin JJ, Mur JM, Wild P, et al. Oral cavity and laryngeal cancers among man-made mineral fiber production workers. Scand J Work Environ Health 1986;12:27-31. Rothman KJ, Cann CI, Flanders D, et al. Epidemiology of laryngeal cancer. Epidemiol Rev 1980;2:195-209. Flanders WD, Cann CI, Rothman KJ. Work-related risk factors for laryngeal cancer. Am J Epidemiol 1984;119:23-32. Olsen J, Sabroe S. Occupational causes of laryngeal cancer. J Epidemiol Community Health 1984;38:117-21. Olsen J, Sabroe S, Lajer M. Welding and cancer of the larynx: a case-control study. Eur J Cancer Clin Oncol 1984;20: 639-43. Zagraniski RT, Kelsey J, Walter SD. Occupational risk factors for laryngeal carcinoma: Connecticut, 1975-1980. Am J Epidemiol 1986;124:67-76. Brownson RC, Chang JC. Exposure to alcohol and tobacco and the risk of laryngeal cancer. Arch Environ Health 1987;42: 192-6. Puntoni R, Goldsmith DF, Vercelli M, et al. A cohort study of workers employed in a refractory brick plant. Scand J Work Environ Health 1987;2:162. International Agency for Research on Cancer. Monographs on the evaluation of carcinogenic risk of chemicals to humans. Vol 37. Tobacco habits other than smoking. Lyon: IARC, 1985. Waldron HA. The carcinogenicity of oil mist. Br J Cancer
1975;32:256-7.
22 Simonato L, Saracci R. Cancer occupational. In: Encyclopedia of occupational health and safety. Vol 1. Geneva: International Labour Office, 1983:369-75.
Accepted 13 November 1989
